Proceedings of the 2000 DOE Hydrogen Program Review
ADVANCED INTERNAL COMBUSTION ENGINE RESEARCH
Peter Van Blarigan
Sandia National Laboratories
Livermore, CA 94550
In this manuscript, research on hydrogen internal combustion engines is discussed. The
objective of this project is to provide a means of renewable hydrogen based fuel utilization. The
development of a high efficiency, low emissions electrical generator will lead to establishing a
path for renewable hydrogen based fuel utilization. A full-scale prototype will be produced in
collaboration with commercial manufacturers.
The electrical generator is based on developed internal combustion engine technology. It is able
to operate on many hydrogen-containing fuels. The efficiency and emissions are comparable to
fuel cells (50% fuel to electricity, ~ 0 NOx). This electrical generator is applicable to both
stationary power and hybrid vehicles. It also allows specific markets to utilize hydrogen
economically and painlessly.
Two motivators for the use of hydrogen as an energy carrier today are: 1) to provide a transition
strategy from hydrocarbon fuels to a carbonless society and 2) to enable renewable energy
sources. The first motivation requires a little discussion while the second one is self-evident.
The most common and cost effective way to produce hydrogen today is the reformation of
hydrocarbon fuels, specifically natural gas. Robert Williams discusses the cost and viability of
natural gas reformation with CO2 sequestration as a cost-effective way to reduce our annual CO2
emission levels. He argues that if a hydrogen economy was in place then the additional cost of
natural gas reformation and subsequent CO2 sequestration is minimal (Williams 1996).
Decarbonization of fossil fuels with subsequent CO2 sequestration to reduce or eliminate our CO2
atmospheric emissions provides a transition strategy to a renewable, sustainable, carbonless
society. However, this requires hydrogen as an energy carrier.
The objectives of this program for the year 2000 are to continue to design, build, and test the
advanced electrical generator components, research hydrogen based renewable fuels, and
develop industrial partnerships. The rationale behind the continuation of designing, building,
and testing generator components is to produce a research prototype for demonstration in two
years. Similarly, researching hydrogen based renewable fuels will provide utilization
components for the largest possible application. Finally, developing industrial partnerships can
lead to the transfer of technology to the commercial sector as rapidly as possible.
This year work is being done on the linear alternator, two-stroke cycle scavenging system,
electromagnetic/combustion/dynamic modeling, and fuel research. The Sandia alternator design
and prototype will be finished, and the Sandia and Magnequench designs will be tested. Work
on the scavenging system consists of learning to use KIVA-3V, and designing the scavenging
experiment. Ron Moses of Los Alamos National Laboratories is conducting the modeling;
modeling of the alternator is being performed. Hydrogen based renewables, such as biogas and
ammonia, are the fuels being researched. Outside of modeling and research, an industrial
collaboration has been made with Caterpillar and Magnequench International, a major supplier
of rare earth permanent magnet materials. A collaborative research and development agreement
(CRADA) has been arranged with Caterpillar, and Magnequench is designing and supplying a
linear alternator. In addition, the prestigious Harry Lee Van Horning Award presented by the
Society of Automotive Engineers (SAE) was awarded in October 1999 for a paper concerning
homogeneous charge compression ignition (HCCI) with a free piston (SAE 982484).
Electrical generators capable of high conversion efficiencies and extremely low exhaust
emissions will no doubt power advanced hybrid vehicles and stationary power systems. Fuel
cells are generally considered to be ideal devices for these applications where hydrogen or
methane are used as fuel. However, the extensive development of the IC engine, and the
existence of repair and maintenance industries associated with piston engines provide strong
incentives to remain with this technology until fuel cells are proven reliable and cost
competitive. In addition, while the fuel cell enjoys high public relations appeal, it seems possible
that it may not offer significant efficiency advantages relative to an optimized combustion
system. In light of these factors, the capabilities of internal combustion engines have been
In regards to thermodynamic efficiency, the Otto cycle theoretically represents the best option
for an IC engine cycle. This is due to the fact that the fuel energy is converted to heat at constant
volume when the working fluid is at maximum compression. This combustion condition leads to
the highest possible peak temperatures, and thus the highest possible thermal efficiencies.
Edson (1964) analytically investigated the efficiency potential of the ideal Otto cycle using
compression ratios (CR) up to 300:1, where the effects of chemical dissociation, working fluid
thermodynamic properties, and chemical species concentration were included. He found that
even as the compression ratio is increased to 300:1, the thermal efficiency still increases for all
of the fuels investigated. At this extreme operating for instance, the cycle efficiency for
isooctane fuel at stoichiometric ratio is over 80%.
Indeed it appears that no fundamental limit exists to achieving high efficiency from an internal
combustion engine cycle. However, many engineering challenges are involved in approaching
ideal Otto cycle performance in real systems, especially where high compression ratios are
Caris and Nelson (1959) investigated the use of high compression ratios for improving the
thermal efficiency of a production V8 spark ignition engine. They found that operation at
compression ratios above about 17:1 did not continue to improve the thermal efficiency in their
configuration. They concluded that this was due to the problem of non-constant volume
combustion, as time is required to propagate the spark-ignited flame.
In addition to the problem of burn duration, other barriers exist. These include the transfer of
heat energy from the combustion gases to the cylinder walls, as well as the operating difficulties
associated with increased pressure levels for engines configured to compression ratios above
25:1 (Overington and Thring 1981, Muranaka and Ishida 1987). Still, finite burn duration
remains the fundamental challenge to using high compression ratios.
The goal of emissions compliance further restricts the design possibilities for an optimized IC
engine. For example, in order to eliminate the production of nitrogen oxides (NOx), the fuel/air
mixture must be homogeneous and very lean at the time of combustion (Das 1990, Van Blarigan
1995). (It is subsequently possible to use oxidation catalyst technologies to sufficiently control
other regulated emissions such as HC and CO.) Homogeneous operation precludes diesel- type
combustion, and spark-ignition operation on premixed charges tends to limit the operating
compression ratio due to uncontrolled autoignition, or knock. As well, very lean fuel/air
mixtures are difficult, or impossible to spark ignite.
On the other hand, lean charges have more favorable specific heat ratios relative to
stoichiometric mixtures, and this leads to improved cycle thermal efficiencies. Equivalence ratio
is no longer required to be precisely controlled, as is required in conventional stoichiometric
operation when utilizing tree way catalysts. Equivalence ratio is defined here as the ratio of the
actual fuel/air ratio to the stoichiometric ratio.
Homogeneous charge compression ignition combustion could be used to solve the problems of
burn duration and allow ideal Otto cycle operation to be more closely approached. In this
combustion process a homogeneous charge of fuel and air is compression heated to the point of
autoignition. Numerous ignition points throughout the mixture can ensure very rapid
combustion (Onishi et al 1979). Very low equivalence ratios (φ ~ 0.3) can be used since no
flame propagation is required. Further, the useful compression ratio can be increased as higher
temperatures are required to autoignite weak mixtures (Karim and Watson 1971).
HCCI operation is unconventional, but is not new. As early as 1957 Alperstein et al. (1958)
experimented with premixed charges of hexane and air, and n-heptane and air in a Diesel engine.
They found that under certain operating conditions their single cylinder engine would run quite
well in a premixed mode with no fuel injection whatsoever.
In general, HCCI combustion has been shown to be faster than spark ignition or compression
ignition combustion. And much leaner operation is possible than in SI engines, while lower NOx
Most of the HCCI studies to date however, have concentrated on achieving smooth releases of
energy under conventional compression condition (CR ~ 9:1). Crankshaft driven pistons have
been utilized in all of these previous investigations. Because of these operating parameters,
successful HCCI operation has required extensive EGR and/or intake air preheating.
Conventional pressure profiles have resulted (Thring 1989, Najt and Foster 1983).
In order to maximize the efficiency potential of HCCI operation much higher compression ratios
must be used, and a very rapid combustion event must be achieved. Recent work with higher
compression ratios (~21:1) has demonstrated the high efficiency potential of the HCCI process
(Christensen et al 1998, Christensen et al 1997).
In Figure 1, the amount of work attained from a modern 4-stroke heavy duty diesel engine is
shown at a 16.25 : 1 compression ratio. The results show that under ideal Otto cycle conditions
(constant volume combustion), 56% more work is still available. This extreme case of non-ideal
Otto cycle behavior serves to emphasize how much can be gained by approaching constant
Co ns tan t Vo lum e C om b ustio n
D ie sel En gin e
P ressure (Pa)
6 5 6 % More W o rk In Co nstant 10
V olum e C om bustio n Cycle
0.00 01 0.0 01
Vo lum e (m e ters )
Figure 1 – Modern 4-Stroke Heavy Duty Diesel Engine
The free piston linear alternator illustrated in Figure 2 has been designed in hopes of approaching
ideal Otto cycle performance through HCCI operation. In this configuration, high compression
ratios can be used and rapid combustion can be achieved.
Figure 2 – Free piston linear alternator
The linear generator is designed such that electricity is generated directly from the piston’s
oscillating motion, as rare earth permanent magnets fixed to the piston are driven back and forth
through the alternator’s coils. Combustion occurs alternately at each end of the piston and a
modern two-stroke cycle scavenging process is used. The alternator component controls the
piston’s motion, and thus the extent of cylinder gas compression, by efficiently managing the
piston’s kinetic energy through each stroke. Compression of the fuel/air mixture is achieved
inertially and as a result, a mechanically simple, variable compression ratio design is possible
with sophisticated electronic control.
The use of free pistons in internal combustion engines has been investigated for quite some time.
In the 1950’s, experiments were conducted with free piston engines in automotive applications.
In these early designs, the engine was used as a gasifier for a single stage turbine (Underwood
1957, Klotsch 1959). More recent developments have integrated hydraulic pumps into the
engine’s design (Baruah 1988, Achten 1994).
Several advantages have been noted for free piston IC engines. First, the compression ratio of
the engine is variable; this is dependent mainly on the engine’s operating conditions (e.g., fuel
type, equivalence ratio, temperature, etc.). As a result, the desired compression ratio can be
achieved through modification of the operating parameters, as opposed to changes in the
An additional benefit is that the mechanical friction can be reduced relative to crankshaft driven
geometries since there is only one moving engine part and no piston side loads. Also,
combustion seems to be faster than in conventional slider-crank configurations. Further, the
unique piston dynamics (characteristically non-sinusoidal) seem to improve the engine’s fuel
economy and NOx emissions by limiting the time that the combustion gases spend at top dead
center (TDC) (thereby reducing engine heat transfer and limiting the NOx kinetics). Finally, one
researcher (Braun 1973) reports that the cylinder/piston/ring wear characteristics are superior to
slider/crank configurations by a factor of 4.
The combination of the HCCI combustion process and the free piston geometry is expected to
result in significant improvements in the engine’s thermal efficiency and its exhaust emissions.
The following advantages should be found:
1. For a given maximum piston velocity, the free piston arrangement is capable of
achieving a desired compression ratio more quickly than a crankshaft driven piston
configuration. This point is illustrated in Figure 3 where the piston position profiles
of both configurations are plotted. The reduced compression time should result in
higher compression of the premixed charge before the onset of autoignition.
V = 2000 cm/s
-5 Free Piston Motion
0 0.005 0.01 0.015 0.02 0.025 0.03
Figure 3 – Piston position vs. time
2. High compression ratio operation is better suited to the free piston engine since the
piston develops compression inertially, and as such there are no bearings or kinematic
constraints that must survive high cylinder pressures or the high rates of pressure
increase (shock). The use of low equivalence ratios in the HCCI application should
further reduce the possibility of combustion chamber surface destruction (Lee and
Schaefer 1983, Maly et al 1990).
3. The free piston design is more capable of supporting the low IMEP levels inherent in
low equivalence ratio operation due to the reduction in mechanical friction.
Integration of the linear alternator into the free piston geometry provides further benefits to the
generator design. In this arrangement mechanical losses in the system are dramatically reduced
since there is essentially one moving part, and this allows engine operation at a more or less
constant piston speed. These points aid in the generator design, and further improve the fuel-to-
electricity generation efficiency of the device.
The linear alternator itself is based on technology developed for brushless DC motors. This class
of motors is characterized by high efficiency and high power density, typically 96% efficiency
and 1 hp per pound density. Put simply, the rotary configuration is unrolled until flat, then rolled
back up perpendicular to the first unrolling to arrive at the linear configuration. Relative to the
rotary geometry the linear device is approximately 30% heavier due to not all of the coils being
driven at the same time. Efficiency will be comparable.
Inherent in the configuration selected is the need to scavenge the exhaust gases out of the
cylinder and replace them with fresh fuel/air charge while the piston is down at the bottom of the
cylinder. This requirement is due to the need to have trapped gases in the cylinder to act as a
spring, as well as to provide the next combustion event.
Conventional 2-stroke cycle engines have developed a reputation for low fuel efficiency and
high hydrocarbon emissions due to short-circuiting of the inlet fuel/air mixture directly to the
exhaust port. The typical 2-stroke application stresses power density over efficiency and
emissions – chain saws, weed whackers, marine outboard motors. These devices must operate
over a wide speed and power range.
In this case the requirements are quite different. The speed of the free piston oscillation is
essentially fixed. Power is varied by modification of the equivalence ratio, not the quantity of
gas delivered. Power density is not a driving requirement. As a result, the design of this system
can be optimized within tight constraints utilizing computational fluid dynamics and
experimental gas dynamics techniques.
Experimental Results - FY 2000
Figure 4 shows the results of experimental combustion studies completed with hydrogen. In this
investigation, a single-stroke rapid compression-expansion machine has been used to
compression ignite hydrogen. Hydrogen is the fastest burning fuel out of all the fuels tested.
The high rate of combustion does approach constant volume combustion. Figure 3 shows a
typical logarithmic P/V diagram for hydrogen combustion at top dead center at 33:1 compression
ratio. The piston has, for all practical purposes, not moved during the combustion event. In the
free piston configuration high pressure-rise rates can be handled without difficulty since there are
no load bearing linkages, as in crankshaft-driven engines. Additionally, operation at equivalence
ratios less than 0.5 reduces the need to consider piston erosion, or other physical damage (Maly
et al. 1990).
Logarithmic Pressure Volume Diagram
Compression Ratio : 33:1, Indicated Thermal Efficiency : 57%
Equivalence Ratio : 0.319, Initial Temperature : 24C
-10.5 -10 -9.5 -9 -8.5 -8 -7.5 -7 -6.5
Figure 4 - Hydrogen Combustion
Figure 5 shows the free piston generator again. The overall length of the generator is 76
centimeters, its specific power is 800 watts per kilogram, and it has a power density of 800 watts
per liter. Hydrogen based renewable fuels such as bio-gas (low BTU producer gas H2-CH4-CO),
ammonia (NH3), methanol (CH4O), and/or hydrogen (H2) can be used directly.
Bio - Gas
Figure 5 – Free Piston Generator
The alternator consists of moving rare earth permanent magnets and stationary output coils and
stator laminations. The design is similar to a conventional rotary brushless DC generator.
Figure 6 shows the magnetic flux path for the linear alternator. It can be seen that the flux
through the coils changes direction as the permanent magnet assembly moves down the
alternator core. This changing flux induces current in the coils.
Two parallel paths are being pursued to develop the linear alternator. An alternator is being built
and tested in house. As a design tool, we are utilizing a two dimensional finite element computer
code to solve Maxwell’s equations of electromagnetism. The code, called FLUX2D, is produced
by MagSoft Corporation. We have investigated various design configurations, and have
optimized a design with respect to maximizing efficiency and minimizing size. In parallel
Magnequench, a commercial development partner, is also designing and fabricating an
alternator. Both alternator designs are being fabricated and will be tested under full design
output conditions on a Sandia designed Caterpillar engine based tester. The tester will measure
both power output and mechanical to electrical conversion efficiency.
Figure 6 – Alternator Design
Figure 7 – Magnequench Linear Alternator Stator Assembly
Magnequench has delivered three stator assemblies to Sandia, one of which is shown in Figure 7.
Also shown in Figure 7 are a short and a long magnet ring. These magnets are pressed from
neodymium-iron-boron rare earth material and magnetized in the radial direction. Sandia will
assemble the Magnequench supplied magnets to the moving part back iron and provide linear
bearing supports. One assembly will then be returned to Magnequench for their own testing.
12 inches Coil Output
Power Output : 40kW Oriented Grain Silicon steel Laminations
Efficiency : 96% (1600 each)
Weight 60 pounds
Figure 8 – Sandia Linear Alternator Design
Figure 8 shows a cut away of the Sandia alternator design. The power output of the linear
alternator is 40 kW, and has an efficiency of 96%. The Magnequench design is very similar; the
differences are primarily in the coil configuration, magnet fabrication and stator material. The
Sandia magnet assembly is fabricated from 10 degree arc magnet segments, which are
magnetized in a linear direction.
The Sandia stator is an assembly of 1600 laminations punched from anisotropic oriented grain
silicon steel. Each lamination has a small angle ground so the assembly stacks into a cylinder.
The Magnequench stator material is pressed iron powder in an adhesive matrix.
The Magnequench coils consist of a single row winding of flat wire. The Sandia coils contain 78
turns of square cross section wire. The Magnequench coils must be connected in moving groups
of five as the magnet assembly moves in the stator. The Sandia design isolates each coil from
the other coils with a Wheatstone bridge. This has the advantage of not requiring an active
magnet assembly following switching network.
Stator Laminations Angle Grinding Operation
NbFeB Coil Winding
Magnet Segments Fixture
Figure 9 – Linear Alternator Design Components
Some of the components and fabrication tools for the Sandia linear alternator are shown in
In preparation for designing the generator control algorithms a comprehensive mathematical
model of the entire physical system is required. One of the challenging aspects of the system
model is the electromagnetic performance of the linear alternator. The finite element model
(FLUX2D) utilized in the design process is too cumbersome for use in a real time system model.
To circumvent this situation we have contracted with Ron Moses of Los Alamos National
Laboratory to derive a simplified alternator model. Ron is an expert in electromagnetic systems
and excited to be part of our project team. We intend to have Ron develop, with our input, a
total system model capable of predicting control system response.
Two Stroke-Cycle Scavenging System Design
Conventional two-stroke cycle engines are designed to maximize power density at the expense of
efficiency and emissions. They also must operate over a wide speed and power range.
Our design intent is to maximize efficiency while minimizing emissions at a narrow power
output operating condition. As a result, the configuration of the scavenging ports and operating
pressures is likely to be unique to this design.
Our approach is to utilize KIVA-3V to design the scavenging system and to validate the KIVA-
3V predictions at selected conditions. Towards this goal we have designed an add-on
scavenging experiment for our free piston combustion test facility. Figure 10 shows the
scavenging experiment on the upper left side connected to the existing combustion experiment
on the lower right side. The experiment will reproduce combustion cylinder pressure and
temperature conditions immediately prior to scavenging port opening and replicate piston motion
during one scavenging cycle. By measuring gases in the cylinder and in the exhaust collector we
will be able to discern trapping efficiency and scavenging efficiency during realistic operating
Figures 11 and 12 show the KIVA-3V modeling results for one particular configuration being
investigated. We are striving to design a loop scavenged flow system due to the simplicity it
possesses. We are also designing a uniflow system with an exhaust valve in the cylinder head.
This approach is thought to be a more confident solution from a fluid mechanics viewpoint.
Figure 10 – Scavenging Experiment
Figure 11 – KIVA-3V flow pictorial
Figure 12 – KIVA-3V grid
Hydrogen Based Renewable Fuels
One of the unique characteristics of HCCI combustion with a free piston is the ability to combust
extremely lean mixtures. In the field of gasification of biomass the simplest approach is to
combust the material in an oxygen-starved environment. The resultant gas is a mixture of
hydrogen, carbon monoxide, carbon dioxide, methane, and nitrogen. The mixture is too lean for
utilization in spark-ignition engines and requires a pilot diesel fuel injection when fumigated into
a diesel engine.
Figure 13 shows the results of combustion of a typical low BTU producer gas as would be
produced from a crude gasifier as would be found in a developing country. The formulation was
kindly supplied by William Hauserman of Hauserman Associates. The results indicate excellent
performance in the free piston experiment. In fact, this lean mixture is ideal for achieving the
NOx control our concept is based upon.
CR : 24:1, Tini : 82C
Fuel Mole Fractions Air/Fuel Mix Mole Fractions
H : 13.3% H : 3.2%
CH : 3.7% CH : 0.8%
CO : 21.5% CO : 5.1% φ : 0.368
CO : 13.4% CO : 3.2%
N2 : 48.1% N : 71.8%
O : 16.1%
1 10 100
Log Volume(in )
Figure 13 – Bio-gas Ideal Otto Cycle Performance
The two most challenging aspects of widespread hydrogen application are the storage of
hydrogen for mobile applications and the distribution infrastructure. In the United States
approximately 1,000,000 farms have access to anhydrous ammonia. The distribution
infrastructure already exists to deliver approximately 8 billion pounds of anhydrous ammonia to
these farms for direct use as a nitrogen soil supplement. The farmers are already handling
anhydrous ammonia and could easily use it as a fuel for their farm equipment if an efficient
utilization device was available.
In Figure 14, the combustion of ammonia exhibits ideal Otto cycle performance in our free
piston combustion experiment, and produces conversion efficiencies comparable to hydrogen
(see Figure 15). Ammonia is an ideal hydrogen based renewable fuel to use in our free piston
generator for several reasons. Ammonia is widely available. 35,000,000,000 pounds of
anhydrous ammonia are produced in the United States per year. Ammonia contains no carbon,
and can be easily made from hydrogen or natural gas.
CR : 48:1, Tini : 63C
1 10 100
Log Volume(in )
Figure 14 – Ammonia Ideal Otto Cycle Performance
Anhydrous ammonia is stored in the same manner as propane, as a liquid under approximately
100 pounds per square inch vapor pressure at room temperature. If released into the atmosphere,
ammonia’s density is lighter than that of air and thus dissipates rapidly. In addition, because of
its characteristic smell the nose easily detects it in concentrations as low as 5 ppm. Finally,
ammonia has such a narrow flammability range that it is generally considered non-flammable
Ammonia is comparable to gasoline as a fuel for combustion engines. Three gallons of ammonia
is equivalent to one gallon of gasoline in energy content. In other terms, 2.35 pounds of
ammonia is equivalent to one pound of gasoline in energy content. Cost wise in 1998, bulk
ammonia was $1.13 per gallon gasoline equivalent.
In using ammonia as a fuel, ammonia and air would enter the free piston generator through the
intake port. After combustion, any generated NOx emissions can be readily reduced by reaction
with ammonia over a zeolite according to one of the following two reactions:
4NO + 4NH3 + O2 ! 4N2 + 6H2O
6NO2 + 8NH3 ! 7N2 + 12H2O
Indicated Thermal Efficiency vs Compression Ratio
Indicated Thermal Efficiency(%)
20 40 60 80 100 120
Figure 15 – Ammonia Combustion
As previously discussed Magnequench International, Incorporated is supplying a linear alternator
at no cost in order to develop new applications for rare earth permanent magnets. In addition,
Caterpillar Corporation is entering into a cooperative research and development agreement
(CRADA) with our group. The purpose of the collaboration is for CAT to share their free piston
lubrication and sealing technology with Sandia while in return applying our linear alternator
technology to their free piston hydraulic pump program.
Both of these collaborations will ease the transfer of this exciting new technology to the
Plans for the 2001 fiscal year include completing the two-stroke scavenging system design,
developing a comprehensive system model, designing a prototype starting system, investigating
alternative funding, and quantifying performance of both alternator designs. The principal
objectives are to select a prototype scavenging system, obtain a predictive model of electrical
and mechanical components, select a starting system, and collaborate with industrial partners in
pursuing other funding.
Achten, P. A. J. 1994. “A Review of Free Piston Engine Concepts,” SAE Paper 941776.
Alperstein, M., Swim, W. B. and Schweitzer, P. H. 1958. “Fumigation Kills Smoke – Improves
Diesel Performance,” SAE Transactions, vol. 66, pp.574 – 588.
Baruah, P. C. 1988. “A Free Piston Engine Hydraulic Pump for an Automotive Propulsion
System,” SAE Paper 880658.
Braun, A. T. and Schweitzer, P. H. 1973. “The Braun Linear Engine,” SAE Paper 730185.
Caris, D. F. and Nelson, E. E. 1959. “A New Look at High Compression Engines,” SAE
Transactions, vol. 67, pp. 112-124.
Christensen, M., Johansson, B. and Einewall, P. 1997. “Homogeneous Charge Compression
Ignition (HCCI) Using Isooctane, Ethanol, and Natural Gas – A Comparison With Spark
Ignition Operation,” SAE Paper 972874.
Christensen, M., Johansson, B., Amneus, P. and Mauss, F. 1998. “Supercharged Homogeneous
Charge Compression Ignition,” SAE Paper 980787.
Das, L. M. 1990. “Hydrogen Engines: A View of the Past and a Look Into the Future,”
International Journal of Hydrogen Energy, vol. 15, no. 6, pp. 425 – 443.
Edson, M. H. 1964. “The Influence of Compression Ratio and Dissociation on Ideal Otto Cycle
Engine Thermal Efficiency,’Digital Calculations of Engine Cycles, SAE Prog. in
Technology, vol. 7, pp. 49-64.
Karim, G.A. and Watson, H.C. 1971. “Experimental and Computational Considerations of the
Compression Ignition of Homogeneous Fuel-Oxidant Mixtures,” SAE Paper 710133.
Klotsch, P. 1959. “Ford Free-Piston Engine Development,” SAE Paper 590045.
Lee, W. and Schaefer, H. J. 1983. “Analysis of Local Pressures, Surface Temperatures and
Engine Damages under Knock Conditions,” SAE Transactions, vol. 92, section 2, pp. 511
Maly, R. R., Klein, R., Peters, N. and Konig, G. 1990. “Theoretical and Experimental
Investigation of Knock Induced Surface Destruction,” SAE Transactions, vol. 99, section
3, pp. 99 – 137.
Muranaka, Y. T. and Ishida, T. 1987. “Factors Limiting the Improvement in Thermal Efficiency
of S.I. Engine at Higher Compression Ratio,” SAE Transactions, vol. 96, section 4, pp.
526 – 536.
Najt, P. M. and Foster, D. E. 1983. “Compression – Ignited Homogeneous Charge
Combustion,” SAE Paper 830264
Onishi, S., Jo, S. H., Shoda, K., Jo, P. D. and Kato, S. 1979. “Active Thermo– Atmospheric
Combustion (ATAC) – A New Combustion Process for Internal Combustion Engines,”
SAE Paper 790501.
Overington, M. T. and Thring, R. H. 1981. “Gasoline Engine Combustion – Turbulence and the
Combustion Chamber,” SAE Paper 810017.
Thring, R. H. 1989. “Homogeneous-Charge Compression-Ignition Engines,” SAE Paper
Underwood, A. F. 1957. “The GMR 4-4 ‘Hyprex’ Engine – A Concept of the Free-Piston
Engine for Automotive Use,” SAE Paper 570032.
Van Blarigan, P., and Green R. 1995. “NOx Emission Data Verified in a Hydrogen Fueled
Engine,” Combustion Research Facility News,” vol.17, no.4, January/February.
Williams, Robert. January 1996. “Fuel Decarbonization for Fuel Cell Applications and
Sequestration of the Separated CO2,” PU/CEES Report No. 295.